2024 Best Paper Award

The Editorial Board of the ASCE Journal of Construction Engineering and Management is pleased to announce the recipients of the 2024 Best Paper Award. Hamed Alikhani, Chau Le, H. David Jeong, and Ivan Damnjanovic are coauthors of an article entitled “Sequential Machine Learning for Activity Sequence Prediction from Daily Work Report Data,” which appeared in the September 2023 issue, https://doi.org/10.1061/JCEMD4.COENG-13165.

Reasonable estimations of project duration are needed by contracting agencies before the advertising date so contract time can be set and initial project plans can be made. Such estimates require good predictions of the sequence of critical activities, activity durations, and how much activities may overlap. Relevant personnel often find the forgoing to be very challenging. This article describes the development of a machine learning model that can receive a part of a project schedule and predict the next part, including the actual sequence and overlap, with 94.4% reliability.

This contribution is particularly remarkable because the input data are daily work reports that are ubiquitously available for contracting agencies but rarely used as a data source for machine learning models. The research commenced developing as-built schedules of critical activities from the daily work reports of 720 Montana Department of Transportation projects. The authors describe in detail how the schedules were factorized and input into a long short-term memory recurrent neural network machine learning model. The model recognizes patterns in the duration, sequence, and overlap of activities after it has been trained on factorized input. Given the start of a schedule, the model finishes the schedule and predicts the duration. Additionally, a workaround is provided in case the model encounters an input vector on which it was not trained so that a reasonable approximation of the schedule can be provided. Eighty percent of the data was used to train the model, and 20% was used to evaluate its reliability.

This article is compelling because the authors have addressed a vexing problem for contracting agencies with the novel use of commonly available data (daily work reports) and a machine learning model. The Editorial Board congratulates the authors on this achievement.

This article was selected from a field of 188 technical papers, 16 case studies, and 5 state-of-the-art reviews that were published from July 2023 through June 2024 in the Journal of Construction Engineering and Management. For this awards cycle, the Awards Committee included the following Editorial Board members: Dr. Young Hoon Kwak, Dr. Ximming Li, Dr. Young Bai, Dr. Mehmet E. Ozbek, Dr. Tripp Shealy, Dr. Eddie Rojas, Dr. Abbas Rashidi, Dr. SangHyun Lee, Dr. Seungjun Ahn, Dr. Kunhee Choi, and Dr. Ioannis Brilakis, with Dr. Charles T. Jahren serving as the chair. Special thanks are due to Dr. Min Liu, Dr. Hexu Liu, and Dr. Fei Dai for providing additional reviews for this and other ASCE awards. Thanks also to the other Editorial Board members, authors, reviewers, and readers who help make this journal a success and this award meaningful.